I have been using the Computer Graphics Lab facilities to assist in the design and synthesis of thyroid hormone receptor antagonists. Triiodothyronine (T3), derived from iodinated tyrosine synthesized in the thyroid gland, plays a critical role in differentiation and development in the fetus and functions as a general metabolic regulator in the adult. An excess or deficiency results in either over or understimulated metabolic activity, respectively, with a considerable negative impact on quality of life. While many thyroid agonists are known, no antagonists have been found to date. Thus the standard treatment for hyperthyroidism consists of surgical removal of the thyroid gland or its destruction with radioactive iodine, followed by a lifelong regime of synthetic thyroid hormone. The availability of thyroid antagonizing drugs could spare the thyroid gland in those afflicted with hyperthyroidism while restoring normal metabolic activity. The design of our putative antagonists has been guided by the principles of structure-based drug design. The receptor for the thyroid hormone is a member of the nuclear receptor superfamily, containing modular ligand and DNA binding domains and functioning as a small-molecule ligand induced transcription factor. The crystal structure of the cocrystal of the ligand binding domain complexed with ligand was recently solved by Robert Fletterick's group at UCSF (1). The structure offers a number of clues regarding the nature of transcriptional activation associated with ligand binding. The carboxy terminal helix (helix 12) has been dubbed the "activation domain" as alterations of its sequence have very detrimental effects on the ability to activate transcription. This region lies adjacent to the ligand binding site and it has been postulated that it undergoes a dramatic conformational shift associated with ligand binding. Our design strategy is to make molecules that retain the critical binding determinants of the native ligand, while being sufficiently different in the right way such that there is a different conformational shift with binding that results in an inactive liganded receptor. The activation domain is rich in phenylalanine residues adjacent to the ligand binding cavity; it is possible that derivatizing a thyroid antagonist with a large hydrophobic moiety, such as another aromatic ring, could have the desired effect. Molecular modeling of the crystal structure using Midas on the CGL workstations has provided some guidance in the positioning of additional sidechains on the native ligand. Syntheses are underway for a series of potential antagonists. Once they are tested, the design of a second generation will commence.